• These data are, of course, backwards-looking. We aim to look forwards, and plan for learning & success.
• Meta-analyses standardise metrics so that comparisons can be made across methods and with larger sample sizes.
  Although we cannot be sure of the strength of any single study (often small sample sizes, low power), meta-analyses allow for stronger overall evaluation of the impact of interventions, by allowing for huge sample sizes.

• As Hattie states, almost all actions we take in schools cause learning.
• The hinge-point of d=0.4 is a mean average of all the things that were measured. The average class, doing average things.
• If we are going to put time, effort & resources into planning educational interventions, we want to aim for higher-impact learning (d>0.6). This could be:
  • Building on practices that are supported with strong data already
  • Investigating practices that we value but that might not rank highly, looking at where the weaknesses lie and how we can improve them for a greater learning outcome.
  • Shining a light on our own tool drawer and 'spring cleaning' old habits that might not be working.
• We can use these data as launching-points for further inquiry and action in our own school contexts.
• Although metrics may be based on standardised/ terminal assessment data, there is validity in using these data to think about interventions we take in our own school settings:
  • On what basis are we making decisions?
  • How do we know the impact of our actions - what are the metrics of success?
  • How can we take action on what we learn about our own practices?

Blank Barometer

High-Impact Strategies

Hattie Speaks: Good & Bad Teaching